Proximity to the president can also have a cost within the ranks of the intelligence community, especially if the DNI is not a professional intelligence officer. Like any other group of professionals, intelligence officers prefer to be directed by one of their own, someone who understands them, who shares their values and cultures and who shares some of their experiences. Remember that only three DCIs were professional intelligence officers (Richard Helms, William Colby, Robert Gates) and two had wartime intelligence experience (Allen Dulles and William Casey). The other DCIs tended to be treated skeptically at first by the intelligence community or, more specifically, by the CIA, with some gaining acceptance and others not. Therefore, a DCI who was seen as being too close to the policy makers and was also not a career intelligence officer would be seen as perhaps being more suspect by the rank and file. The same may run true for the DNIs, whose only legal requirement for the job is “extensive national security experience.” The added liability for the DNI is separation from all intelligence agencies, including the CIA. Again, much will depend on the nature of the DNI’s relationship with the president and how DNIs conduct themselves vis-a-vis the rest of the intelligence community.
The intelligence community also wants to be kept informed about policy directions and preferences. Although this would seem obligatory if the intelligence community is expected to provide relevant analysis, it does not always happen. All too often, policy makers do not keep intelligence abreast, either by design or omission. Such behavior not only makes the role of intelligence more difficult but also can lead to resentment that may be played out in other ways.
Another difference between the two groups is that of outlook. As a senior intelligence officer observed, policy makers tend to be optimists. They approach problems with the belief that they can solve them. After all, this is the reason they have gone into government. Intelligence officers are skeptics. Their training teaches them to question and to doubt. Although they may see an optimistic outcome to a given situation, they also see the potential pessimistic outcomes and likely feel compelled to analyze them as potential outcomes.
A revealing indication of the potential costs of the difference in outlook emerged in 2004, when relations between the Bush administration and the CIA deteriorated seriously. Differences over the progress being made in containing the insurgency in Iraq appear to have been the main stimulus. Leaks of intelligence analyses, which some White House officials characterized as being written by “pessimists, naysayers, and handwringers,” exacerbated the problem. At one point, President Bush said the CIA “was just guessing” about potential outcomes in Iraq, a remark that some intelligence officers found demeaning. It became customary to say that the CIA and the White House were “at war.” The fact that the exchange took place in the middle of a presidential election undoubtedly added to the tension. Indeed, the relationship deteriorated to the point where the acting DCI, John McLaughlin, felt it necessary to go to President Bush and assure him that the CIA was not covertly supporting Democratic nominee Sen. John Kerry, Mass., in the election.
Several lessons are derived from this byplay. First, war or warlike situations—especially those that may be inconclusive—tend to increase the overall tension, as can be easily understood. Second, in such circumstances both parties can forget the nature of their relationship, although this is probably a greater problem for the policy makers. The combination of uncertainty and casualties, with the attendant political costs, raise the policy makers’ anxiety. Third, the leadership of the intelligence community understands that it can never win this sort of struggle with policy makers and therefore will seek to avoid such confrontations. The professional ethos and training of senior intelligence officials work to preclude such an outcome. Even if their analyses prove to be correct, the costs to their relationship with senior policy officials would be so great as to result in a Pyrrhic victory. This does not mean that analysts should temper their views or to hedge what they write but that intelligence officers are unlikely to engage in gratuitous and overt hostility to policy makers.
Finally, the policy makers’ expectation of support from the permanent bureaucracy extends to the intelligence community. But they may be seeking intelligence that supports known policy preferences, thus running the risk of politicization. Politicization can also work in the other direction. The intelligence officer’s desire to be listened to (Kent’s second wish) may lead to analysis that is meant to please the policy makers, either consciously or unwttingly. In either case, the desire for a good working relationship can directly undermine the desired objectivity of intelligence. This aspect of the relationship has probably been exacerbated by the increasing practice of Congress levying requests for national intelligence estimates (NIEs) that are essentially progress reports on the war on terror or the situation in Iraq and then also requiring that the Key Judgments (KJs) of these NIEs be declassified and published. Congress is entirely within its right to request NIEs, although these progress report estimates do appear to have political agendas behind them. Publication of the KJs certainly increases the likelihood that the estimates will be used by one or both sides in the political debate. It also increases the likelihood that either the president or Congress or both will assume that unpalatable judgments were written to please opponents in the debate. In October 2007, DNI Mike McConnell decided that NIEs would not be made public any longer because of his concerns about the effect this had on the quality of the analysis. However, as noted, he reversed this decision seven weeks later in the case of the Iran nuclear NIE and allowed the KJs to be published in declassified form.
THE INTELLIGENCE PROCESS: POLICY AND INTELLIGENCE
The differences between the policy and intelligence communities—and the potential for tension—appear at each stage of the intelligence process.
REQUIREMENTS. Requirements are not abstract concepts. They are the policy makers’ agenda. All policy makers have certain areas on which they must concentrate as well as others on which they would like to concentrate. Some areas are of little or no interest to them but require their attention either occasionally or regularly. This mixture of preferences is important in forming the agenda and thus the requirements. For example, Secretary of State James A. Baker III ( 1989-1992) was clear, on taking office, that he was not going to spend a lot of time on the Middle East. His decision was not based on a view that the region was unimportant but that he was unlikely to achieve much in the Middle East and therefore his time would be better spent elsewhere. Senior subordinates could handle the Middle East. Iraq’s invasion of Kuwait undermined his choice. Ironically, the war also helped lead to the Madrid conference—presided over by Secretary Baker—at which Israel and its Arab foes met together for the first time.
The intelligence community wants guidance on the priorities of the agenda so that its collection and reporting can be as helpful as possible. At the same time, the community tends to understand that it rarely has the luxury of ignoring a region or issue entirely, even if it is not high on the agenda because of the global coverage requirement. Sooner or later, one region or issue is likely to blow up. That said, the intelligence community regularly makes resource choices that lead to some regions or issues receiving little attention.
The degree to which each administration formally communicates its requirements—versus relying on the intelligence community to know which issues matter—also varies. President Clinton was willing to go through a formal requirements exercise only once in eight years. Under President George W. Bush the requirements are reviewed every six months. No matter which method is used or the frequency of formal requirements, the intelligence community is held responsible for ensuring that it has collection and analytical resources on the most important issues. Policy makers also tend to expect that the intelligence community anticipates the emergence of new issues. After all, isn’t this one of the main functions of intelligence? The answer is not a firm yes if you take into account the fact that surprises occur: assassinations, coups, elections, reversals of policy. Not everything can be anticipated.
The difference in the approaches of policy makers and intelligence officials to requirements is not played out entirely in formulating the requirements themselves, but in the ensuing phases of the intelligence process.
COLLECTION. Policy makers tend to be divorced from the details of collection unless they involve political sensitivities. In such cases, policy makers can have direct and dramatic effects.
(See
box,
“Policy Makers and Intelligence Collection.
”) Their practical concerns lie, first, in the budget, as collection is one of the major intelligence costs, particularly technical intelligence. But policy makers also tend to assume, incorrectly, that everything is being covered, at least at some minimal level. Thus, when one of the low-priority issues explodes, they expect that a certain low level of collection and on-the-shelf intelligence already exists and that collection can be quickly increased. Both assumptions may be strikingly false. Collection priority decisions tend to be zero-sum games, and not all collection assets are easily fungible.
The intelligence community would rather collect more than less, although intelligence officials recognize that they cannot cover everything and hope to get policy makers to concur in the areas left uncovered. Still, collection is the bedrock of intelligence. But when policy makers place limits on collection, the intelligence community obeys, even if its preference is to collect. Like the policy makers, intelligence officials are aware of the costs of collection, but they cannot spend more on collection than the policy makers (the president and Congress) are willing to allocate. The customary practice is for the policy community to set budgetary limits on collection resources that are lower than the intelligence community would like.
POLICY MAKERS AND INTELLIGENCE COLLECTION
In several instances, policy makers have intervened in intelligence collection for political reasons.
In Cuba, at the onset of the missile crisis in 1962, Secretary of State Dean Rusk opposed sending U-2s over the island because a Chinese Nationalist U-2 had recently been shot down over China and an Air Force U-2 had accidentally violated Soviet air space in Siberia. The need for imagery of possible Soviet missile sites in Cuba was great, but Rusk had other-also legitimate-concerns about avoiding further provocation.
In Iran, several successive U.S. administrations imposed limits on intelligence collection. Basically, intelligence officers were not allowed to have contact with those in the souks (markets and bazaars) who were opposed to the shah, because the shah’s regime would be offended Instead, U.S. intelligence had to rely on the shah’s secret police, Savak, which had an institutional interest in denying that any opposition existed. Thus, as the shah’s regime unraveled in 1978-1979, policy makers denied U.S. intelligence the sources and contacts it needed to better analyze the situation or to influence the opposition
Again, regarding Cuba, President Jimmy Carter unilaterally suspended U-2 flights as a gesture to improve bilateral relations. Carter came to regret his decision in 1980, when he faced the possibility that a Soviet combat brigade was in Cuba and he required better intelligence on the issue.
Finally, the intelligence community has a greater understanding, as would be expected, of the limits of collection at any given time. Intelligence officials know that they are not collecting everything. They make decisions on a regular basis to exclude certain regions or issues. In their own budget requests, intelligence officials also determine how much of the collection to process and exploit, which is always far less than is collected. The intelligence community sees no reasons to convey these facts to the policy makers. At one level, doing so is unnecessary. A region not receiving much collection allocation may stay quiet, which is the bet that the intelligence managers are making. At another level, it may undermine their relationship with policy makers. Why arouse concerns about collection coverage over an issue that is not expected to be a significant priority? Their choices can lead to even worse relations should one of the regions suddenly become a concern and collection be found wanting.
ANALYSIS. Policy makers want information that enables them to make an informed decision, but they do not come to this part of the process as blank slates or wholly objective observers. Already in favor of certain policies and outcomes, they would like to see intelligence that supports their preferences. Again, this is not necessarily as crass as it sounds. Policy makers only naturally prefer intelligence that enables them to go where they want. This attitude becomes problematic only when they ignore intelligence that is compelling but contrary to their preferences.
INTELLIGENCE UNCERTAINTIES AND POLICY
In 1987 U 5.-Soviet negotiations were drawing to a close on the Intermediate Nuclear Forces (INF) Treaty. The US intelligence community had three methods for estimating the number of Soviet INF missiles that had been produced-all of which had to be accounted for and destroyed. Meanwhile, any final number given by the Soviets would be suspect.
Each of the three major intelligence agencies advocated its methodology and its number as the one that should go forward But the senior intelligence officer responsible for the issue decided, correctly, that all three numbers had to go to President Ronald Reagan. Some agency representatives argued that this was simply pusillanimous hedging. But the intelligence officer argued that the president had to be aware of the intelligence uncertainties and the possible range of missile numbers before he signed the treaty. That was the right answer, instead of choosing, perhaps arbitrarily, among the methodologies.